I would like to run the Kalman smoother from the R package imputeTS to impute the missing values of several univariate time series. From the literature it seems that the initialization of the first value might have significant effects on the inference results. But from the documentation of the R package imputeTS it unfortunately does not become clear what the default initialization of the Kalman smoother in imputeTS is. I would appreciate any hint on this.
Thanks in advance, Marco
